The CEA welcomes 1,600 doctoral PhD students to its laboratories each year.
Thesis
Home / Post Doctorat / Multi-Agent Negotiation for Collaborative Resource Placement in Distributed Cloud Networks
Multi-Agent Negotiation for Collaborative Resource Placement in Distributed Cloud Networks
New computing paradigms, circuits and technologies, incl. quantumTechnological challenges
Abstract
This research project aims to design a decentralized and autonomous resource management system for heterogeneous cloud networks. Building on the shift toward distributed architectures driven by concerns over data sovereignty and performance, the project seeks to move beyond the traditional centralized control plane model used in Kubernetes. Each organization involved in a federation of clusters would be represented by an intelligent agent, capable of negotiating resource placement according to its own objectives while preserving data confidentiality. The interaction among these agents is modeled as a multi-agent game, where incentive mechanisms are designed to reach a mutually beneficial equilibrium. The project plans to formalize the problem, adapt multi-agent reinforcement learning methods to the challenges of distributed settings (such as fault tolerance and asynchronous communication), and develop a functional implementation in Rust. In doing so, it lays the groundwork for a new paradigm of collaboration among cloud service providers.
Laboratory
Département Ingénierie Logiciels et Systèmes (LIST)
LICIA (DILS)
Lab.systèmes d’information de confiance, intelligents et auto-organisants
Nous utilisons des cookies pour vous garantir la meilleure expérience sur notre site web. Si vous continuez à utiliser ce site, nous supposerons que vous en êtes satisfait.OKNonPolitique de confidentialité